by bacterial action, but it is lost largely by gaseous diffusion and only 

 in very wet weather by leaching. Carbon dioxide was therefore 

 determined simultaneously with nitrates, and the curves show a 

 marked similarity except that the increases in nitrate came later. 

 Thus we may conclude that the curves are in the main production 

 curves. 



In winter the curves follow the temperature pretty closely, but 

 through the rest of the year they follow the rainfall and to a less 

 extent the moisture. There is evidence that the dissolved oxygen 

 in the rain may be a factor of importance particularly as it ensures 

 renewal of the dissolved atmosphere. After a heavy rainfall the 

 carbon dioxide in the soil air begins to increase followed later by an 

 increase in nitrates. A certain number of bacterial counts were 

 made, and these showed that a rise in bacterial numbers preceded 

 the autumn increase in carbon dioxide. 



A series of field experiments has been started to ascertain how 

 these spring and autumn periods of biochemical activity may best 

 be utilised for crop production. 



The other reaction, the loss of gaseous nitrogen from the soil, 

 is under investigation in the laboratory, but we anticipate that our 

 main guidance will come from the parallel investigation of a manure 

 heap which is being made simultaneously by Mr. E. H. Richards, 

 the Hon. Rupert Guinness Research Chemist. These experiments 

 have indicated a hitherto unsuspected source of loss in the heap and, 

 fortunately, a way of avoiding it. Mr. Richards finds that nitrates 

 are formed on the outside of the heap, but readily decompose if they 

 are washed inside. Watering a heap sufficiently to effect this but 

 not enough to cause leaching was found to increase the loss con- 

 siderably ; on the other hand shelter from rain, together with 

 compacting, reduced the loss practically to zero over a period of 

 three months' storage. 



The technical value of this work is very considerable, but of 

 even greater significance is its application in elucidating the loss of 

 gaseous nitrogen from soil. In at least one direction a soil particle 

 containing organic matter resembles a manure heap : it is surrounded 

 by a free atmosphere containing some 20 per cent, of oxygen so that 

 nitrate production goes on at its surface ; it also has a dissolved 

 atmosphere devoid of oxygen and therefore supplying one of the 

 essential conditions for denitrification. This resemblance furnishes 

 a working hypothesis which is now being developed. 



The activity of the micro-organisms of the soil is dependent on 

 two other factors besides those already mentioned. A source of 

 energy is obviously indispensable : this is provided by the residues 

 of plants and is, indeed, one of the most interesting of the inter- 

 relationships between plants and soil micro-organisms. The other 

 factor is the presence in the soil of sufficient basic material which in 

 practice always means a sufficiency of calcium carbonate. 



The precise effect of this substance has still to be determined, 

 and pending the completion of Mr. Horton's investigations it is safest 

 to follow the practical man and speak of soils as "sour" when calcium 

 carbonate is lacking. "Sour" soils are commonly infertile. The 

 method of overcoming the trouble is to add either lime or chalk. 

 Dr. Hutchinson and Mr. IVIacLennan find, however, that lime differs 

 fundamentally from chalk in its action. Lime is a partial sterilising 



